Adjust logging to avoid expensive CosmosException.getMessage() diagnostics json serialization#49321
Open
jaxpod70 wants to merge 2 commits into
Open
Adjust logging to avoid expensive CosmosException.getMessage() diagnostics json serialization#49321jaxpod70 wants to merge 2 commits into
jaxpod70 wants to merge 2 commits into
Conversation
github-actions
Bot
added
Community Contribution
Contributor
|
Thank you for your contribution @jaxpod70! We will review the pull request and get back to you soon. |
Author
|
@microsoft-github-policy-service agree company="Microsoft" |
Contributor
There was a problem hiding this comment.
Pull request overview
Note
Copilot was unable to run its full agentic suite in this review.
Updates change feed partition processor logging when handling CosmosException, aiming to reduce WARN log verbosity while still allowing full exception details at DEBUG.
Changes:
- Switched WARN logs to parameterized logging and appended
clientException.getShortMessage(). - Added guarded DEBUG logs that include the full exception/stack trace.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 5 comments.
| File | Description |
|---|---|
| sdk/cosmos/azure-cosmos/src/main/java/com/azure/cosmos/implementation/changefeed/pkversion/PartitionProcessorImpl.java | Refines CosmosException logging in pk-version processor (WARN summary + DEBUG stack trace). |
| sdk/cosmos/azure-cosmos/src/main/java/com/azure/cosmos/implementation/changefeed/epkversion/PartitionProcessorImpl.java | Refines CosmosException logging in epk-version processor (WARN summary + DEBUG stack trace). |
Comment on lines
+211
to
+216
| logger.warn( | ||
| "CosmosException: partition {} from thread {} with owner {} {}", | ||
| this.lease.getLeaseToken(), | ||
| Thread.currentThread().getId(), | ||
| this.lease.getOwner(), | ||
| clientException.getShortMessage()); |
Comment on lines
+218
to
+221
| logger.debug( | ||
| "CosmosException: partition" + this.lease.getLeaseToken() | ||
| + "from thread " + Thread.currentThread().getId() + " with owner " + this.lease.getOwner(), | ||
| clientException); |
| + " from thread " + Thread.currentThread().getId() + " with owner " + this.lease.getOwner(), | ||
| clientException); | ||
| logger.warn( | ||
| "CosmosException: partition {} from thread {} with owner {} {}", |
Comment on lines
+201
to
+205
| "Lease with token {}: CosmosException was thrown from thread {} for lease with owner {} {}", | ||
| this.lease.getLeaseToken(), | ||
| Thread.currentThread().getId(), | ||
| this.lease.getOwner(), | ||
| clientException.getShortMessage()); |
Comment on lines
+207
to
+210
| logger.debug( | ||
| "Lease with token " + this.lease.getLeaseToken() + ": CosmosException was thrown from thread " + | ||
| Thread.currentThread().getId() + " for lease with owner " + this.lease.getOwner(), | ||
| clientException); |
|
|
||
| CosmosException clientException = (CosmosException) throwable; | ||
| logger.warn( | ||
| "Lease with token " + this.lease.getLeaseToken() + ": CosmosException was thrown from thread " + |
Member
There was a problem hiding this comment.
Using ShortMessage is fine for possibly noisy logs - see Exceptions.isCommonlyExpectedExceptionPossiblyCausingNoisyLogs for an example - but for unexpected errors the full diagnostics are crucial for being able to debug - so, please use the Exceptions.isCommonlyExpectedExceptionPossiblyCausingNoisyLogs to decide whether to use full or short message
FabianMeiswinkel
requested changes
Jun 1, 2026
Member
FabianMeiswinkel
left a comment
FabianMeiswinkel
left a comment
There was a problem hiding this comment.
Please honor Exceptions.isCommonlyExpectedExceptionPossiblyCausingNoisyLogs
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Fixes excessive log line sizes in the ChangeFeedProcessor's partition processing loop caused by CosmosException.getMessage() serializing the full CosmosDiagnostics payload (request timelines, region contacts, retry history, metadata) into JSON when logged at WARN level.
Under heavy DB connection load, this produces multi-MB log lines (30+ MB observed) on every transient error (429 throttles, timeouts, connectivity issues) across all leased partitions concurrently. This causes log ingestion pipeline throttling/rejection, increased storage costs, and memory pressure from repeated serialization in a hot loop.
Fix: Split the existing log statement into two tiers in both epkversion.PartitionProcessorImpl and pkversion.PartitionProcessorImpl:
This preserves observability for operators at default log levels while eliminating the excessive payload.
Related Issue
49320
All SDK Contribution checklist:
General Guidelines and Best Practices (https://github.com/Azure/azure-sdk-for-java/blob/main/CONTRIBUTING.md#developer-guide)
Testing Guidelines (https://github.com/Azure/azure-sdk-for-java/blob/main/CONTRIBUTING.md#building-and-unit-testing)